We provide a comprehensive and updated systematic review and meta-analysis of the association between air pollution exposure and depression, searching PubMed, Embase, and Web of Sciences for relevant articles published up to May 2021, and eventually including 39 studies. Meta-analyses were performed separately according to pollutant type [particulate matter with diameter ≤10 μm (PM₁₀) and ≤2.5 μm (PM_2.5), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), ozone (O₃), and carbon monoxide (CO)] and exposure duration [short- (<30 days) and long-term (≥30 days)]. Test for homogeneity based on Cochran's Q and I² statistics were calculated and the restricted maximum likelihood (REML) random effect model was applied. We assessed overall quality of pooled estimates, influence of single studies on the meta-analytic estimates, sources of between-study heterogeneity, and publication bias. We observed an increased risk of depression associated with long-term exposure to PM_2.5 (relative risk: 1.074, 95% confidence interval: 1.021-1.129) and NO₂ (1.037, 1.011-1.064), and with short-term exposure to PM₁₀ (1.009, 1.006-1.012), PM_2.5 (1.009, 1.007-1.011), NO₂ (1.022, 1.012-1.033), SO₂ (1.024, 1.010-1.037), O₃ (1.011, 0.997-1.026), and CO (1.062, 1.020-1.105). The publication bias affecting half of the investigated associations and the high heterogeneity characterizing most of the meta-analytic estimates partly prevent to draw very firm conclusions. On the other hand, the coherence of all the estimates after excluding single studies in the sensitivity analysis supports the soundness of our results. This especially applies to the association between PM_2.5 and depression, strengthened by the absence of heterogeneity and of relevant publication bias in both long- and short-term exposure studies. Should further investigations be designed, they should involve large sample sizes, well-defined diagnostic criteria for depression, and thorough control of potential confounding factors. Finally, studies dedicated to the comprehension of the mechanisms underlying the association between air pollution and depression remain necessary.

This study was designed to investigate and compare demographic and clinical features with specific emphasis on age at onset, age at first treatment and, in particular, on duration of untreated illness (DUI), in patients with different mood and anxiety disorders. Study sample included 729 outpatients with the following diagnoses: major depressive disorder (n=181), bipolar disorder type I (BD I, n=115) and II (BD II, n=186), generalized anxiety disorder (n=100), panic disorder (n=96), and obsessive–compulsive disorder (n=51). Main demographic and clinical variables of the sample were compared among the diagnostic groups using one-way analysis of variance or χ² tests. The diagnostic groups showed significant differences in relation to age at onset and age at first pharmacological treatment and in relation to latency to treatment. In particular, patients with major depressive disorder showed the shortest DUI (39.08 months), whereas patient with BD II showed the longest DUI (97.2 months) in comparison with the other groups. Within the group with anxiety disorders (F=7.512, P<0.001), patients with panic disorder showed the shortest DUI (44.35 months), whereas patients with obsessive–compulsive disorder showed the longest DUI (90.57 months). The present findings suggest that patients with different mood and anxiety disorders show significant differences in terms of age at onset, age at first treatment and, consequently, DUI, which potentially reflect different reasons influencing treatment delay.